Benchmark reports
IPB Reports
Public IPB reports are scoped evidence artifacts, not surprise leaderboard drops. The current report domain is Enterprise Copilot Safety v0.2. Frontier and open-weight report branches will publish after release gates are complete.
Enterprise Copilot Safety v0.2
Public release is scheduled for July 22, 2026. Reports will include scoped findings, charts, caveats, vendor-response status where applicable, and selected public-safe examples. Live corpus generation, held-out challenge sets, and future test material remain closed.
Scheduled public release
Frontier Model Reports
The first frontier report set is scoped to IPB Enterprise Copilot Safety v0.2. Public release is gated by evidence validation, private vendor preview, challenge review, public-safe redaction, caveat review, and release approval.
Topline Protocol Score
July 22, 2026
Publishing July 22, 2026
Correctness vs. Stability
July 22, 2026
Publishing July 22, 2026
In preparation
Open-Weight Model Reports
The open-weight branch will use the same ECS v0.2 methodology and public disclosure boundaries, with additional reproducibility context for downloadable model configurations where appropriate.
Topline Protocol Score
July 22, 2026
Publishing July 22, 2026
Correctness vs. Stability
July 22, 2026
Publishing July 22, 2026
Report non-claims
- IPB is not a universal intelligence ranking.
- IPB is not a claim that a model is globally safe.
- IPB is not certification.
- IPB does not replace legal, regulatory, security, medical, financial, or compliance review.
- IPB results are scoped to the declared domain, protocol version, corpus version, model/system identity, and runtime settings.
- Stable behavior is not automatically good behavior; stable-wrong behavior is a failure.
- Public samples do not disclose future test material.